Given a list of winning probabilities, this algorithm backs out a relative ability of each horse in the race.
Applicable Scenarios and Problems
Let's say you have market implied information about horse ability. This algorithm converts market prices into a relative rating which can be subsequently used to analyze performance in past and future races.
Describe the input fields for your algorithm. For example:
|input||List of winning probabilities|
Winning probabilities will be normalized to sum to 1 if that is not already the case.
|ability||A numerical ability score|
This algorithm assumes horse performance is skew-normally distributed (horses are more likely to perform very badly than extremely well). The location parameter of each horse is adjusted until the probabilities of each horse winning match the supplied vector of probabilities.
More details and link to paper will be supplied soon.